Face Generation¶

In this project, you'll use generative adversarial networks to generate new images of faces.

Get the Data¶

You'll be using two datasets in this project:

MNIST
CelebA

Since the celebA dataset is complex and you're doing GANs in a project for the first time, we want you to test your neural network on MNIST before CelebA. Running the GANs on MNIST will allow you to see how well your model trains sooner.

If you're using FloydHub, set data_dir to "/input" and use the FloydHub data ID "R5KrjnANiKVhLWAkpXhNBe".

#data_dir = './data'

# FloydHub - Use with data ID "R5KrjnANiKVhLWAkpXhNBe"
data_dir = '/input'


"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import helper

helper.download_extract('mnist', data_dir)
helper.download_extract('celeba', data_dir)

Downloading mnist: 9.92MB [00:11, 895KB/s]                             
Extracting mnist: 100%|██████████| 60.0K/60.0K [00:12<00:00, 4.90KFile/s]
Downloading celeba: 1.44GB [00:22, 64.4MB/s]

Extracting celeba...

Explore the Data¶

MNIST¶

As you're aware, the MNIST dataset contains images of handwritten digits. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
%matplotlib inline
import os
from glob import glob
from matplotlib import pyplot

mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'mnist/*.jpg'))[:show_n_images], 28, 28, 'L')
pyplot.imshow(helper.images_square_grid(mnist_images, 'L'), cmap='gray')

<matplotlib.image.AxesImage at 0x7fde645b44a8>

CelebA¶

The CelebFaces Attributes Dataset (CelebA) dataset contains over 200,000 celebrity images with annotations. Since you're going to be generating faces, you won't need the annotations. You can view the first number of examples by changing show_n_images.

show_n_images = 25

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
mnist_images = helper.get_batch(glob(os.path.join(data_dir, 'img_align_celeba/*.jpg'))[:show_n_images], 28, 28, 'RGB')
pyplot.imshow(helper.images_square_grid(mnist_images, 'RGB'))

<matplotlib.image.AxesImage at 0x7fde5eaac278>

Preprocess the Data¶

Since the project's main focus is on building the GANs, we'll preprocess the data for you. The values of the MNIST and CelebA dataset will be in the range of -0.5 to 0.5 of 28x28 dimensional images. The CelebA images will be cropped to remove parts of the image that don't include a face, then resized down to 28x28.

The MNIST images are black and white images with a single color channel while the CelebA images have 3 color channels (RGB color channel).

Build the Neural Network¶

You'll build the components necessary to build a GANs by implementing the following functions below:

model_inputs
discriminator
generator
model_loss
model_opt
train

Check the Version of TensorFlow and Access to GPU¶

This will check to make sure you have the correct version of TensorFlow and access to a GPU

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
from distutils.version import LooseVersion
import warnings
import tensorflow as tf

# Check TensorFlow Version
assert LooseVersion(tf.__version__) >= LooseVersion('1.0'), 'Please use TensorFlow version 1.0 or newer.  You are using {}'.format(tf.__version__)
print('TensorFlow Version: {}'.format(tf.__version__))

# Check for a GPU
if not tf.test.gpu_device_name():
    warnings.warn('No GPU found. Please use a GPU to train your neural network.')
else:
    print('Default GPU Device: {}'.format(tf.test.gpu_device_name()))

TensorFlow Version: 1.3.1
Default GPU Device: /gpu:0

Input¶

Implement the model_inputs function to create TF Placeholders for the Neural Network. It should create the following placeholders:

Real input images placeholder with rank 4 using image_width, image_height, and image_channels.
Z input placeholder with rank 2 using z_dim.
Learning rate placeholder with rank 0.

Return the placeholders in the following the tuple (tensor of real input images, tensor of z data)

import problem_unittests as tests

def model_inputs(image_width, image_height, image_channels, z_dim):
    """
    Create the model inputs
    :param image_width: The input image width
    :param image_height: The input image height
    :param image_channels: The number of image channels
    :param z_dim: The dimension of Z
    :return: Tuple of (tensor of real input images, tensor of z data, learning rate)
    """
    real_inputs = tf.placeholder(tf.float32, (None, image_width, image_height, image_channels), name = "real_inputs")
    z_inputs = tf.placeholder(tf.float32, (None, z_dim), name = "z_inputs")
    learning_rate = tf.placeholder(tf.float32, name = "learning_rate")
    
    return real_inputs, z_inputs, learning_rate


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_inputs(model_inputs)

ERROR:tensorflow:==================================
Object was never used (type <class 'tensorflow.python.framework.ops.Operation'>):
<tf.Operation 'assert_rank_2/Assert/Assert' type=Assert>
If you want to mark it as used call its "mark_used()" method.
It was originally created here:
['File "/usr/local/lib/python3.6/runpy.py", line 193, in _run_module_as_main\n    "__main__", mod_spec)', 'File "/usr/local/lib/python3.6/runpy.py", line 85, in _run_code\n    exec(code, run_globals)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel_launcher.py", line 16, in <module>\n    app.launch_new_instance()', 'File "/usr/local/lib/python3.6/site-packages/traitlets/config/application.py", line 658, in launch_instance\n    app.start()', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/kernelapp.py", line 477, in start\n    ioloop.IOLoop.instance().start()', 'File "/usr/local/lib/python3.6/site-packages/zmq/eventloop/ioloop.py", line 177, in start\n    super(ZMQIOLoop, self).start()', 'File "/usr/local/lib/python3.6/site-packages/tornado/ioloop.py", line 888, in start\n    handler_func(fd_obj, events)', 'File "/usr/local/lib/python3.6/site-packages/tornado/stack_context.py", line 277, in null_wrapper\n    return fn(*args, **kwargs)', 'File "/usr/local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 440, in _handle_events\n    self._handle_recv()', 'File "/usr/local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 472, in _handle_recv\n    self._run_callback(callback, msg)', 'File "/usr/local/lib/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 414, in _run_callback\n    callback(*args, **kwargs)', 'File "/usr/local/lib/python3.6/site-packages/tornado/stack_context.py", line 277, in null_wrapper\n    return fn(*args, **kwargs)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 283, in dispatcher\n    return self.dispatch_shell(stream, msg)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 235, in dispatch_shell\n    handler(stream, idents, msg)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 399, in execute_request\n    user_expressions, allow_stdin)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/ipkernel.py", line 196, in do_execute\n    res = shell.run_cell(code, store_history=store_history, silent=silent)', 'File "/usr/local/lib/python3.6/site-packages/ipykernel/zmqshell.py", line 533, in run_cell\n    return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)', 'File "/usr/local/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2698, in run_cell\n    interactivity=interactivity, compiler=compiler, result=result)', 'File "/usr/local/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2808, in run_ast_nodes\n    if self.run_code(code, result):', 'File "/usr/local/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2862, in run_code\n    exec(code_obj, self.user_global_ns, self.user_ns)', 'File "<ipython-input-5-7b5e62e0ea17>", line 22, in <module>\n    tests.test_model_inputs(model_inputs)', 'File "/output/problem_unittests.py", line 12, in func_wrapper\n    result = func(*args)', 'File "/output/problem_unittests.py", line 68, in test_model_inputs\n    _check_input(learn_rate, [], \'Learning Rate\')', 'File "/output/problem_unittests.py", line 34, in _check_input\n    _assert_tensor_shape(tensor, shape, \'Real Input\')', 'File "/output/problem_unittests.py", line 20, in _assert_tensor_shape\n    assert tf.assert_rank(tensor, len(shape), message=\'{} has wrong rank\'.format(display_name))', 'File "/usr/local/lib/python3.6/site-packages/tensorflow/python/ops/check_ops.py", line 617, in assert_rank\n    dynamic_condition, data, summarize)', 'File "/usr/local/lib/python3.6/site-packages/tensorflow/python/ops/check_ops.py", line 571, in _assert_rank_condition\n    return control_flow_ops.Assert(condition, data, summarize=summarize)', 'File "/usr/local/lib/python3.6/site-packages/tensorflow/python/util/tf_should_use.py", line 175, in wrapped\n    return _add_should_use_warning(fn(*args, **kwargs))', 'File "/usr/local/lib/python3.6/site-packages/tensorflow/python/util/tf_should_use.py", line 144, in _add_should_use_warning\n    wrapped = TFShouldUseWarningWrapper(x)', 'File "/usr/local/lib/python3.6/site-packages/tensorflow/python/util/tf_should_use.py", line 101, in __init__\n    stack = [s.strip() for s in traceback.format_stack()]']
==================================
Tests Passed

Discriminator¶

Implement discriminator to create a discriminator neural network that discriminates on images. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "discriminator" to allow the variables to be reused. The function should return a tuple of (tensor output of the discriminator, tensor logits of the discriminator).

def discriminator(images, reuse=False):
    """
    Create the discriminator network
    :param images: Tensor of input image(s)
    :param reuse: Boolean if the weights should be reused
    :return: Tuple of (tensor output of the discriminator, tensor logits of the discriminator)
    """
    keep_prob = 0.8
    alpha = 0.3
    
    with tf.variable_scope("discriminator", reuse = reuse):
        x1 = tf.layers.conv2d(images, 64, 5, strides = 2, padding = "SAME", kernel_initializer=tf.contrib.layers.xavier_initializer())
        x1 = tf.maximum(alpha * x1, x1)
        x1 = tf.nn.dropout(x1, keep_prob)

        x2 = tf.layers.conv2d(x1, 128, 5, strides = 2, padding = "SAME", kernel_initializer=tf.contrib.layers.xavier_initializer())
        x2 = tf.layers.batch_normalization(x2, training = True)
        x2 = tf.maximum(alpha * x2, x2)
        x2 = tf.nn.dropout(x2, keep_prob)
        
        x3 = tf.layers.conv2d(x2, 256, 5, strides = 2, padding = "SAME", kernel_initializer=tf.contrib.layers.xavier_initializer())
        x3 = tf.layers.batch_normalization(x3, training = True)
        x3 = tf.maximum(alpha * x3, x3)
        x3 = tf.nn.dropout(x3, keep_prob)
        
        flat = tf.reshape(x3, (-1, 4* 4* 256))
        logits = tf.layers.dense(flat, 1, kernel_initializer=tf.contrib.layers.xavier_initializer())
        logits = tf.nn.dropout(logits, keep_prob)
        output = tf.sigmoid(logits)
        
    return output, logits


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_discriminator(discriminator, tf)

Tests Passed

Generator¶

Implement generator to generate an image using z. This function should be able to reuse the variables in the neural network. Use tf.variable_scope with a scope name of "generator" to allow the variables to be reused. The function should return the generated 28 x 28 x out_channel_dim images.

def generator(z, out_channel_dim, is_train=True):
    """
    Create the generator network
    :param z: Input z
    :param out_channel_dim: The number of channels in the output image
    :param is_train: Boolean if generator is being used for training
    :return: The tensor output of the generator
    """
    alpha = 0.3
    keep_prob = 0.5
    with tf.variable_scope("generator", reuse = not is_train):
        x1 = tf.layers.dense(z, 2 * 2 * 512)
        x1 = tf.reshape(x1, (-1, 2, 2, 512))
        x1 = tf.layers.batch_normalization(x1, training = is_train)
        x1 = tf.maximum(alpha * x1, x1)
        x1 = tf.nn.dropout(x1, keep_prob)
        
        x2 = tf.layers.conv2d_transpose(x1, 256, 5, strides = 2, padding = "VALID")
        x2 = tf.layers.batch_normalization(x2, training = is_train)
        x2 = tf.maximum(alpha * x2, x2)
        x2 = tf.nn.dropout(x2, keep_prob)

        x3 = tf.layers.conv2d_transpose(x2, 128, 5, strides = 2, padding = "SAME")
        x3 = tf.layers.batch_normalization(x3, training = is_train)
        x3 = tf.maximum(alpha * x3, x3)        
        x3 = tf.nn.dropout(x3, keep_prob)
        
        logits = tf.layers.conv2d_transpose(x3, out_channel_dim, 5, strides = 2, padding = "SAME")
        output = tf.tanh(logits)
        
    return output


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_generator(generator, tf)

Tests Passed

Loss¶

Implement model_loss to build the GANs for training and calculate the loss. The function should return a tuple of (discriminator loss, generator loss). Use the following functions you implemented:

discriminator(images, reuse=False)
generator(z, out_channel_dim, is_train=True)

def model_loss(input_real, input_z, out_channel_dim):
    """
    Get the loss for the discriminator and generator
    :param input_real: Images from the real dataset
    :param input_z: Z input
    :param out_channel_dim: The number of channels in the output image
    :return: A tuple of (discriminator loss, generator loss)
    """
    
    g_model = generator(input_z, out_channel_dim)
    d_model_real, d_logits_real = discriminator(input_real)
    d_model_fake, d_logits_fake = discriminator(g_model, reuse = True)
    
    d_loss_real = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits = d_logits_real, labels = tf.ones_like(d_model_real) * 0.9))
    d_loss_fake = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits = d_logits_fake, labels = tf.zeros_like(d_model_fake)))
    g_loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits = d_logits_fake, labels = tf.ones_like(d_model_fake) * 0.9))
    
    d_loss = d_loss_real + d_loss_fake
    
    return d_loss, g_loss


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_loss(model_loss)

Tests Passed

Optimization¶

Implement model_opt to create the optimization operations for the GANs. Use tf.trainable_variables to get all the trainable variables. Filter the variables with names that are in the discriminator and generator scope names. The function should return a tuple of (discriminator training operation, generator training operation).

def model_opt(d_loss, g_loss, learning_rate, beta1):
    """
    Get optimization operations
    :param d_loss: Discriminator loss Tensor
    :param g_loss: Generator loss Tensor
    :param learning_rate: Learning Rate Placeholder
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :return: A tuple of (discriminator training operation, generator training operation)
    """
    t_vars = tf.trainable_variables()
    d_vars = [var for var in t_vars if var.name.startswith("discriminator")]
    g_vars = [var for var in t_vars if var.name.startswith("generator")]
    
    with tf.control_dependencies(tf.get_collection(tf.GraphKeys.UPDATE_OPS)):
        d_train_opt = tf.train.AdamOptimizer(learning_rate, beta1 = beta1).minimize(d_loss, var_list = d_vars)
        g_train_opt = tf.train.AdamOptimizer(learning_rate, beta1 = beta1).minimize(g_loss, var_list = g_vars)

    return d_train_opt, g_train_opt


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
tests.test_model_opt(model_opt, tf)

Tests Passed

Neural Network Training¶

Show Output¶

Use this function to show the current output of the generator during training. It will help you determine how well the GANs is training.

"""
DON'T MODIFY ANYTHING IN THIS CELL
"""
import numpy as np

def show_generator_output(sess, n_images, input_z, out_channel_dim, image_mode):
    """
    Show example output for the generator
    :param sess: TensorFlow session
    :param n_images: Number of Images to display
    :param input_z: Input Z Tensor
    :param out_channel_dim: The number of channels in the output image
    :param image_mode: The mode to use for images ("RGB" or "L")
    """
    cmap = None if image_mode == 'RGB' else 'gray'
    z_dim = input_z.get_shape().as_list()[-1]
    example_z = np.random.uniform(-1, 1, size=[n_images, z_dim])

    samples = sess.run(
        generator(input_z, out_channel_dim, False),
        feed_dict={input_z: example_z})

    images_grid = helper.images_square_grid(samples, image_mode)
    pyplot.imshow(images_grid, cmap=cmap)
    pyplot.show()

Train¶

Implement train to build and train the GANs. Use the following functions you implemented:

model_inputs(image_width, image_height, image_channels, z_dim)
model_loss(input_real, input_z, out_channel_dim)
model_opt(d_loss, g_loss, learning_rate, beta1)

Use the show_generator_output to show generator output while you train. Running show_generator_output for every batch will drastically increase training time and increase the size of the notebook. It's recommended to print the generator output every 100 batches.

def train(epoch_count, batch_size, z_dim, learning_rate, beta1, get_batches, data_shape, data_image_mode):
    """
    Train the GAN
    :param epoch_count: Number of epochs
    :param batch_size: Batch Size
    :param z_dim: Z dimension
    :param learning_rate: Learning Rate
    :param beta1: The exponential decay rate for the 1st moment in the optimizer
    :param get_batches: Function to get batches
    :param data_shape: Shape of the data
    :param data_image_mode: The image mode to use for images ("RGB" or "L")
    """
    if(data_image_mode == "L"):
        out_channel_dim = 1
    else:
        out_channel_dim = 3
    
    batches, img_width, img_height, img_depth = data_shape
    input_real, input_z, lr = model_inputs(img_width, img_height, img_depth, z_dim)
    
    d_loss, g_loss = model_loss(input_real, input_z, img_depth)
    d_opt, g_opt = model_opt(d_loss, g_loss, learning_rate, beta1)
    
    steps = 0
    print_every = 10
    show_every = 100
    losses = []
    num_images = 25
    
    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        for epoch_i in range(epoch_count):
            for batch_images in get_batches(batch_size):
                steps += 1
                batch_images *= 2.0
                batch_z = np.random.uniform(-1, 1, (batch_size, z_dim))
                
                _ = sess.run(d_opt, feed_dict = {input_real: batch_images, input_z: batch_z, lr: learning_rate})
                _ = sess.run(g_opt, feed_dict = {input_real: batch_images, input_z: batch_z, lr: learning_rate})
                _ = sess.run(g_opt, feed_dict = {input_real: batch_images, input_z: batch_z, lr: learning_rate})
                
                if steps%print_every == 0:
                    train_loss_d = d_loss.eval({input_z: batch_z, input_real: batch_images})
                    train_loss_g = g_loss.eval({input_z: batch_z})
                    
                    print("Epoch {}/{}...".format(epoch_i+1, epoch_count),
                          "Discriminator Loss: {:.4f}...".format(train_loss_d),
                          "Generator Loss: {:.4f}".format(train_loss_g))
                    losses.append((train_loss_d, train_loss_g))
                
                if steps%show_every == 0:
                    show_generator_output(sess, num_images, input_z, img_depth, data_image_mode)

MNIST¶

Test your GANs architecture on MNIST. After 2 epochs, the GANs should be able to generate images that look like handwritten digits. Make sure the loss of the generator is lower than the loss of the discriminator or close to 0.

batch_size = 32
z_dim = 100
learning_rate = 0.0002
beta1 = 0.2


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 5

mnist_dataset = helper.Dataset('mnist', glob(os.path.join(data_dir, 'mnist/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, mnist_dataset.get_batches,
          mnist_dataset.shape, mnist_dataset.image_mode)

Epoch 1/5... Discriminator Loss: 3.1325... Generator Loss: 0.4632
Epoch 1/5... Discriminator Loss: 2.4425... Generator Loss: 0.4357
Epoch 1/5... Discriminator Loss: 2.0789... Generator Loss: 0.5501
Epoch 1/5... Discriminator Loss: 2.4789... Generator Loss: 0.5417
Epoch 1/5... Discriminator Loss: 1.3451... Generator Loss: 1.1364
Epoch 1/5... Discriminator Loss: 1.5238... Generator Loss: 0.7561
Epoch 1/5... Discriminator Loss: 2.0583... Generator Loss: 0.5986
Epoch 1/5... Discriminator Loss: 1.6008... Generator Loss: 0.8135
Epoch 1/5... Discriminator Loss: 1.6085... Generator Loss: 0.5963
Epoch 1/5... Discriminator Loss: 1.8848... Generator Loss: 1.1945

Epoch 1/5... Discriminator Loss: 1.8152... Generator Loss: 1.1816
Epoch 1/5... Discriminator Loss: 1.7456... Generator Loss: 0.7826
Epoch 1/5... Discriminator Loss: 1.0871... Generator Loss: 1.4259
Epoch 1/5... Discriminator Loss: 1.0717... Generator Loss: 1.6354
Epoch 1/5... Discriminator Loss: 1.1210... Generator Loss: 1.2430
Epoch 1/5... Discriminator Loss: 1.4000... Generator Loss: 0.9158
Epoch 1/5... Discriminator Loss: 1.4496... Generator Loss: 1.7171
Epoch 1/5... Discriminator Loss: 1.0912... Generator Loss: 1.2496
Epoch 1/5... Discriminator Loss: 1.4538... Generator Loss: 1.0268
Epoch 1/5... Discriminator Loss: 1.3916... Generator Loss: 1.3541

Epoch 1/5... Discriminator Loss: 1.5341... Generator Loss: 0.7705
Epoch 1/5... Discriminator Loss: 1.4262... Generator Loss: 1.6787
Epoch 1/5... Discriminator Loss: 1.2739... Generator Loss: 1.3514
Epoch 1/5... Discriminator Loss: 1.2192... Generator Loss: 1.6895
Epoch 1/5... Discriminator Loss: 1.6142... Generator Loss: 0.6219
Epoch 1/5... Discriminator Loss: 1.4403... Generator Loss: 0.7260
Epoch 1/5... Discriminator Loss: 1.8204... Generator Loss: 0.5348
Epoch 1/5... Discriminator Loss: 1.4663... Generator Loss: 0.7915
Epoch 1/5... Discriminator Loss: 1.0692... Generator Loss: 1.4382
Epoch 1/5... Discriminator Loss: 1.3612... Generator Loss: 0.8201

Epoch 1/5... Discriminator Loss: 1.2396... Generator Loss: 0.8683
Epoch 1/5... Discriminator Loss: 1.4284... Generator Loss: 0.7339
Epoch 1/5... Discriminator Loss: 1.2216... Generator Loss: 1.1706
Epoch 1/5... Discriminator Loss: 1.3088... Generator Loss: 0.8274
Epoch 1/5... Discriminator Loss: 1.5975... Generator Loss: 1.4815
Epoch 1/5... Discriminator Loss: 1.2681... Generator Loss: 0.7330
Epoch 1/5... Discriminator Loss: 1.1805... Generator Loss: 1.0815
Epoch 1/5... Discriminator Loss: 1.4635... Generator Loss: 1.4774
Epoch 1/5... Discriminator Loss: 1.3888... Generator Loss: 1.3198
Epoch 1/5... Discriminator Loss: 1.1422... Generator Loss: 1.0630

Epoch 1/5... Discriminator Loss: 1.3569... Generator Loss: 0.8137
Epoch 1/5... Discriminator Loss: 1.1903... Generator Loss: 1.1885
Epoch 1/5... Discriminator Loss: 1.2796... Generator Loss: 0.9722
Epoch 1/5... Discriminator Loss: 1.3171... Generator Loss: 0.8748
Epoch 1/5... Discriminator Loss: 1.3779... Generator Loss: 0.9347
Epoch 1/5... Discriminator Loss: 1.4693... Generator Loss: 0.9798
Epoch 1/5... Discriminator Loss: 1.4297... Generator Loss: 0.7744
Epoch 1/5... Discriminator Loss: 1.2490... Generator Loss: 1.5321
Epoch 1/5... Discriminator Loss: 1.3352... Generator Loss: 0.8136
Epoch 1/5... Discriminator Loss: 1.4492... Generator Loss: 0.5516

Epoch 1/5... Discriminator Loss: 1.1189... Generator Loss: 1.1706
Epoch 1/5... Discriminator Loss: 1.1384... Generator Loss: 0.9046
Epoch 1/5... Discriminator Loss: 1.6104... Generator Loss: 1.9145
Epoch 1/5... Discriminator Loss: 1.2453... Generator Loss: 1.0343
Epoch 1/5... Discriminator Loss: 1.1917... Generator Loss: 1.2151
Epoch 1/5... Discriminator Loss: 1.3021... Generator Loss: 0.9010
Epoch 1/5... Discriminator Loss: 1.1982... Generator Loss: 1.0137
Epoch 1/5... Discriminator Loss: 1.1860... Generator Loss: 1.2259
Epoch 1/5... Discriminator Loss: 1.0330... Generator Loss: 1.0322
Epoch 1/5... Discriminator Loss: 1.5476... Generator Loss: 0.7584

Epoch 1/5... Discriminator Loss: 1.1977... Generator Loss: 0.8426
Epoch 1/5... Discriminator Loss: 1.1480... Generator Loss: 0.7031
Epoch 1/5... Discriminator Loss: 1.2802... Generator Loss: 1.0689
Epoch 1/5... Discriminator Loss: 1.1574... Generator Loss: 1.0179
Epoch 1/5... Discriminator Loss: 0.9867... Generator Loss: 1.1304
Epoch 1/5... Discriminator Loss: 1.2939... Generator Loss: 0.9276
Epoch 1/5... Discriminator Loss: 1.1641... Generator Loss: 1.2643
Epoch 1/5... Discriminator Loss: 1.1897... Generator Loss: 0.9723
Epoch 1/5... Discriminator Loss: 1.2227... Generator Loss: 1.0121
Epoch 1/5... Discriminator Loss: 1.4250... Generator Loss: 1.9670

Epoch 1/5... Discriminator Loss: 1.1741... Generator Loss: 1.2219
Epoch 1/5... Discriminator Loss: 1.0566... Generator Loss: 1.0376
Epoch 1/5... Discriminator Loss: 1.1794... Generator Loss: 0.9112
Epoch 1/5... Discriminator Loss: 1.1398... Generator Loss: 1.9341
Epoch 1/5... Discriminator Loss: 1.0397... Generator Loss: 1.1491
Epoch 1/5... Discriminator Loss: 1.4234... Generator Loss: 0.7200
Epoch 1/5... Discriminator Loss: 1.0363... Generator Loss: 1.4731
Epoch 1/5... Discriminator Loss: 1.2276... Generator Loss: 1.2856
Epoch 1/5... Discriminator Loss: 1.1285... Generator Loss: 1.3204
Epoch 1/5... Discriminator Loss: 1.1498... Generator Loss: 1.1794

Epoch 1/5... Discriminator Loss: 1.2794... Generator Loss: 0.7581
Epoch 1/5... Discriminator Loss: 1.3247... Generator Loss: 1.0884
Epoch 1/5... Discriminator Loss: 1.1640... Generator Loss: 1.0154
Epoch 1/5... Discriminator Loss: 1.4331... Generator Loss: 0.6389
Epoch 1/5... Discriminator Loss: 1.1258... Generator Loss: 1.0097
Epoch 1/5... Discriminator Loss: 1.2464... Generator Loss: 0.9662
Epoch 1/5... Discriminator Loss: 1.3501... Generator Loss: 0.7154
Epoch 1/5... Discriminator Loss: 1.3456... Generator Loss: 1.4808
Epoch 1/5... Discriminator Loss: 1.0955... Generator Loss: 1.2074
Epoch 1/5... Discriminator Loss: 1.0572... Generator Loss: 0.8829

Epoch 1/5... Discriminator Loss: 1.2079... Generator Loss: 2.4270
Epoch 1/5... Discriminator Loss: 1.2042... Generator Loss: 0.7273
Epoch 1/5... Discriminator Loss: 1.3381... Generator Loss: 0.8053
Epoch 1/5... Discriminator Loss: 1.1275... Generator Loss: 1.4474
Epoch 1/5... Discriminator Loss: 1.0495... Generator Loss: 1.1839
Epoch 1/5... Discriminator Loss: 1.2537... Generator Loss: 0.5899
Epoch 1/5... Discriminator Loss: 0.9088... Generator Loss: 1.3840
Epoch 1/5... Discriminator Loss: 1.1884... Generator Loss: 1.8805
Epoch 1/5... Discriminator Loss: 0.9998... Generator Loss: 1.2158
Epoch 1/5... Discriminator Loss: 0.9779... Generator Loss: 1.0929

CelebA¶

Run your GANs on CelebA. It will take around 20 minutes on the average GPU to run one epoch. You can run the whole epoch or stop when it starts to generate realistic faces.

batch_size = 32
z_dim = 100
learning_rate = 0.0002
beta1 = 0.3


"""
DON'T MODIFY ANYTHING IN THIS CELL THAT IS BELOW THIS LINE
"""
epochs = 2

celeba_dataset = helper.Dataset('celeba', glob(os.path.join(data_dir, 'img_align_celeba/*.jpg')))
with tf.Graph().as_default():
    train(epochs, batch_size, z_dim, learning_rate, beta1, celeba_dataset.get_batches,
          celeba_dataset.shape, celeba_dataset.image_mode)

Epoch 1/2... Discriminator Loss: 0.9472... Generator Loss: 2.1614
Epoch 1/2... Discriminator Loss: 1.9095... Generator Loss: 0.6176
Epoch 1/2... Discriminator Loss: 0.9277... Generator Loss: 2.5319
Epoch 1/2... Discriminator Loss: 1.0260... Generator Loss: 2.3705
Epoch 1/2... Discriminator Loss: 1.6402... Generator Loss: 1.2342
Epoch 1/2... Discriminator Loss: 2.1283... Generator Loss: 0.5272
Epoch 1/2... Discriminator Loss: 1.1333... Generator Loss: 4.3287
Epoch 1/2... Discriminator Loss: 1.0682... Generator Loss: 1.7071
Epoch 1/2... Discriminator Loss: 1.0132... Generator Loss: 1.7653
Epoch 1/2... Discriminator Loss: 0.9003... Generator Loss: 1.4157

Epoch 1/2... Discriminator Loss: 1.0535... Generator Loss: 1.3390
Epoch 1/2... Discriminator Loss: 1.0773... Generator Loss: 1.2080
Epoch 1/2... Discriminator Loss: 1.0267... Generator Loss: 2.5822
Epoch 1/2... Discriminator Loss: 0.7797... Generator Loss: 2.0038
Epoch 1/2... Discriminator Loss: 0.9953... Generator Loss: 2.5389
Epoch 1/2... Discriminator Loss: 1.1538... Generator Loss: 1.1471
Epoch 1/2... Discriminator Loss: 0.9660... Generator Loss: 1.1617
Epoch 1/2... Discriminator Loss: 0.8838... Generator Loss: 1.8944
Epoch 1/2... Discriminator Loss: 2.2287... Generator Loss: 0.5555
Epoch 1/2... Discriminator Loss: 0.7805... Generator Loss: 1.5942

Epoch 1/2... Discriminator Loss: 1.0569... Generator Loss: 1.5327
Epoch 1/2... Discriminator Loss: 0.7897... Generator Loss: 1.9522
Epoch 1/2... Discriminator Loss: 1.4510... Generator Loss: 2.9610
Epoch 1/2... Discriminator Loss: 1.2771... Generator Loss: 0.8316
Epoch 1/2... Discriminator Loss: 0.8942... Generator Loss: 2.3350
Epoch 1/2... Discriminator Loss: 1.0723... Generator Loss: 1.2100
Epoch 1/2... Discriminator Loss: 1.1318... Generator Loss: 0.9154
Epoch 1/2... Discriminator Loss: 1.0164... Generator Loss: 2.4290
Epoch 1/2... Discriminator Loss: 1.3082... Generator Loss: 0.8083
Epoch 1/2... Discriminator Loss: 1.2279... Generator Loss: 2.4841

Epoch 1/2... Discriminator Loss: 1.5571... Generator Loss: 0.7173
Epoch 1/2... Discriminator Loss: 0.9456... Generator Loss: 1.6267
Epoch 1/2... Discriminator Loss: 1.9307... Generator Loss: 0.6246
Epoch 1/2... Discriminator Loss: 1.1614... Generator Loss: 1.8207
Epoch 1/2... Discriminator Loss: 1.2405... Generator Loss: 1.0086
Epoch 1/2... Discriminator Loss: 1.7136... Generator Loss: 2.4594
Epoch 1/2... Discriminator Loss: 1.2877... Generator Loss: 1.4629
Epoch 1/2... Discriminator Loss: 1.0784... Generator Loss: 1.3167
Epoch 1/2... Discriminator Loss: 1.0491... Generator Loss: 1.7415
Epoch 1/2... Discriminator Loss: 1.1849... Generator Loss: 0.9127

Epoch 1/2... Discriminator Loss: 1.2020... Generator Loss: 1.5474
Epoch 1/2... Discriminator Loss: 1.1388... Generator Loss: 0.9117
Epoch 1/2... Discriminator Loss: 1.0719... Generator Loss: 1.4641
Epoch 1/2... Discriminator Loss: 1.0842... Generator Loss: 1.8562
Epoch 1/2... Discriminator Loss: 1.0044... Generator Loss: 1.6554
Epoch 1/2... Discriminator Loss: 1.0642... Generator Loss: 1.7611
Epoch 1/2... Discriminator Loss: 1.2821... Generator Loss: 0.7759
Epoch 1/2... Discriminator Loss: 1.3096... Generator Loss: 0.9676
Epoch 1/2... Discriminator Loss: 1.2021... Generator Loss: 1.1383
Epoch 1/2... Discriminator Loss: 1.3438... Generator Loss: 1.3218

Epoch 1/2... Discriminator Loss: 1.7293... Generator Loss: 1.0976
Epoch 1/2... Discriminator Loss: 1.4507... Generator Loss: 0.7895
Epoch 1/2... Discriminator Loss: 1.0449... Generator Loss: 1.1355
Epoch 1/2... Discriminator Loss: 1.1427... Generator Loss: 1.1516
Epoch 1/2... Discriminator Loss: 1.4004... Generator Loss: 0.6183
Epoch 1/2... Discriminator Loss: 1.5494... Generator Loss: 0.7347
Epoch 1/2... Discriminator Loss: 0.8655... Generator Loss: 1.2973
Epoch 1/2... Discriminator Loss: 1.0703... Generator Loss: 1.2049
Epoch 1/2... Discriminator Loss: 1.4005... Generator Loss: 1.0119
Epoch 1/2... Discriminator Loss: 1.6174... Generator Loss: 1.4992

Epoch 1/2... Discriminator Loss: 1.0547... Generator Loss: 1.2576
Epoch 1/2... Discriminator Loss: 1.0442... Generator Loss: 0.9491
Epoch 1/2... Discriminator Loss: 1.0984... Generator Loss: 1.5227
Epoch 1/2... Discriminator Loss: 1.0921... Generator Loss: 1.1717
Epoch 1/2... Discriminator Loss: 1.3876... Generator Loss: 0.9504
Epoch 1/2... Discriminator Loss: 1.2184... Generator Loss: 0.8969
Epoch 1/2... Discriminator Loss: 1.1501... Generator Loss: 1.0237
Epoch 1/2... Discriminator Loss: 1.0875... Generator Loss: 1.0325
Epoch 1/2... Discriminator Loss: 1.1881... Generator Loss: 1.0680
Epoch 1/2... Discriminator Loss: 1.3398... Generator Loss: 0.9459

Epoch 1/2... Discriminator Loss: 1.6082... Generator Loss: 0.7144
Epoch 1/2... Discriminator Loss: 1.1498... Generator Loss: 1.2005
Epoch 1/2... Discriminator Loss: 1.3004... Generator Loss: 1.1448
Epoch 1/2... Discriminator Loss: 1.1800... Generator Loss: 0.8762
Epoch 1/2... Discriminator Loss: 1.3163... Generator Loss: 0.7967
Epoch 1/2... Discriminator Loss: 1.1673... Generator Loss: 1.3496
Epoch 1/2... Discriminator Loss: 1.2274... Generator Loss: 1.0668
Epoch 1/2... Discriminator Loss: 1.2672... Generator Loss: 1.0510
Epoch 1/2... Discriminator Loss: 1.2981... Generator Loss: 0.8676
Epoch 1/2... Discriminator Loss: 1.1060... Generator Loss: 1.0575

Epoch 1/2... Discriminator Loss: 1.4109... Generator Loss: 1.1348
Epoch 1/2... Discriminator Loss: 1.5436... Generator Loss: 1.8239
Epoch 1/2... Discriminator Loss: 1.0077... Generator Loss: 0.9944
Epoch 1/2... Discriminator Loss: 1.2479... Generator Loss: 1.3148
Epoch 1/2... Discriminator Loss: 1.2974... Generator Loss: 1.3898
Epoch 1/2... Discriminator Loss: 1.2983... Generator Loss: 0.8678
Epoch 1/2... Discriminator Loss: 0.9975... Generator Loss: 1.2613
Epoch 1/2... Discriminator Loss: 1.1894... Generator Loss: 1.1487
Epoch 1/2... Discriminator Loss: 1.3728... Generator Loss: 0.9612
Epoch 1/2... Discriminator Loss: 1.1452... Generator Loss: 0.9573

Epoch 1/2... Discriminator Loss: 1.2306... Generator Loss: 0.8492
Epoch 1/2... Discriminator Loss: 1.4209... Generator Loss: 0.9200
Epoch 1/2... Discriminator Loss: 1.1433... Generator Loss: 1.0686
Epoch 1/2... Discriminator Loss: 1.3993... Generator Loss: 1.0352
Epoch 1/2... Discriminator Loss: 1.3890... Generator Loss: 0.8784
Epoch 1/2... Discriminator Loss: 1.2042... Generator Loss: 0.9628
Epoch 1/2... Discriminator Loss: 1.0349... Generator Loss: 0.9408
Epoch 1/2... Discriminator Loss: 1.3272... Generator Loss: 0.7762
Epoch 1/2... Discriminator Loss: 1.3753... Generator Loss: 1.2732
Epoch 1/2... Discriminator Loss: 1.1873... Generator Loss: 0.9561

Submitting This Project¶

When submitting this project, make sure to run all the cells before saving the notebook. Save the notebook file as "dlnd_face_generation.ipynb" and save it as a HTML file under "File" -> "Download as". Include the "helper.py" and "problem_unittests.py" files in your submission.